Research On Multi-information Recognition Of Piezoelectric Film Tactile Sensor

Tactile is a kind of perception ability to obtain the attribute information of objects when people contact with external objects directly.Tactile sensor can provide the attribute information of contact objects and assist the robot to distinguish different objects accurately.At present,tactile signal detection system is widely used in artificial skin and haptic interaction.However,there are still many areas to be improved in real-time acquisition of object attribute information and object multi-information recognition.A novel polyvinylidene fluoride(PVDF)piezoelectric film sensor has the advantages of high flexibility,high sensitivity,stable structure,and piezoelectric and pyroelectric characteristics.It is an ideal component of bionic tactile detection system.Based on the piezoelectric film,a multi-information tactile recognition system with high sensitivity and easy integration is designed,which can recognize the soft and hard,viscoelastic,thermal conductivity and roughness of objects degree and other attribute information.Firstly,this paper summarizes the development status of tactile sensor,and proposes a multi-information recognition system based on PVDF piezoelectric film tactile sensor.PVDF piezoelectric film has high sensitivity and can recognize weak signals,so it can be used as a sensitive unit of tactile sensor.Based on the principle of human finger skin bionics,the structure design of double bump multilayer PVDF piezoelectric film tactile sensor is realized.Because there are some problems in the process of signal transmission,such as impedance matching,noise interference and single-chip microcomputer can only identify positive signal.After the signal acquisition is completed,the data needs to be processed.In order to process the collected signals,the signal conditioning circuits such as charge amplification,low-pass filtering and voltage rise are designed.Then,the sensor comparative test experiment is carried out,which proves that the sensitivity of the proposed double bump multilayer PVDF piezoelectric film tactile sensor is higher than that of the double bump single-layer and double-layer PVDF piezoelectric film tactile sensor.Using the humanoid hand to knock and touch the object,we can obtain the attribute information such as the softness,viscoelasticity and thermal conductivity of the object.Using the time-domain characteristics of the signal,we can preliminarily judge the hardness,viscoelasticity and thermal conductivity of the object.Finally,the paper carries out the experiment of roughness recognition based on PSO-BP algorithm.PCA algorithm is used to reduce the dimension of data.On the premise of ensuring that the most important information in the original sample is not lost,several independent attribute features are selected to identify different roughness samples,which not only ensures the accuracy of the algorithm,but also improves the efficiency of the algorithm and shortens the operation time of the algorithm.The PSO algorithm is used to improve BP neural network,which makes the convergence accuracy of the optimized BP algorithm higher.The results show that the maximum accuracy of PSO-BP algorithm for different roughness samples is 98% and the average recognition accuracy is 94%.In this paper,the multi-information recognition of piezoelectric film tactile sensor is studied,which provides reference for the future research of bionic robot in sensing the external environment and recognizing the tactile information of objects.It has a good application prospect in intelligent robot and prosthetic application.